11 Oct 2007

Researchers in the U.S., Japan, Belgium, and Germany have joined forces to find out how Pen-2, Aph-1, presenilin, and nicastrin do the same. The four proteins are well-known components of γ-secretase, the intramembrane protease that releases amyloid-β from its precursor. That much is clear. What is murkier is how many molecules of each protein go into the protease complex. The answer is one, according to Michael Wolfe, Brigham and Women’s Hospital, Boston, and collaborators in Yasuo Ihara’s lab, University of Tokyo; Bart De Strooper’s lab, KULeuven; and Christian Haass’ lab, Ludwig Maximilians University, Munich.

The one-for-all stoichiometry makes life somewhat simpler than the alternative proposition, which is that the complex contains a presenilin (PS) dimer at the core. Since there are two presenilins (and more than 100 different PS missense mutations identified) a dimer increases the number of different active sites that can be formed.

A definitive answer to this primer/dimer question has remained elusive. The closest researchers have gotten to a structural image of γ-secretase is a 3D electron micrograph, but that did not have sufficient resolution to determine if one or two PS proteins are present (see ARF related news story). Wolfe and colleagues used a biochemical approach. In the October 2 Journal of Biological Chemistry online, first author Toru Sato and colleagues show that immunoprecipitation of the complex always pulls down one presenilin, even if two are present in the cell. For example, when the researchers expressed both myc- and hemagglutinin-tagged PS in PS1/2 double knockout cells, anti-myc never pulled down HA-PS, and anti-HA never pulled down myc-PS. Both antibodies pulled down γ-secretase activity, however, suggesting that active protease can form with a single presenilin. Sato and colleagues obtained similar results when they carried out the experiments in the presence of γ-secretase substrates or inhibitors, just in case dimers may form under conditions that change PS conformation. The results suggest that does not happen, either.

Sato and colleagues then used the knowledge that there is only one PS per complex to quantify the amount of the three other components, arriving at the 1:1:1:1 stoichiometry. This predicts a molecular weight of around 220 kDa. However, native gel electrophoresis indicates the complex is around 300-800 kDa. The authors write that they cannot rule out the possibility that other proteins or substrates contribute to the size of the complex, but they also note that some electrophoretic methods can be “fraught with difficulties.” Another possibility these findings cannot rule out is that two or more γ-secretases form larger complexes in cells, albeit each γ-secretase with a single active PS. What the researchers do rule out is “the suggestion that one PS can affect another PS,” or that “one FAD mutant PS molecule can alter the site of proteolysis by wild-type PS.”—Tom Fagan